Refactor IssueHandler from supervisor to issue_verification_agent

Move the IssueHandler workflow from ymir/supervisor/ into a standalone
BeeAI Framework agent at ymir/agents/issue_verification_agent.py,
following the pattern established by preliminary_testing_agent.

This includes:
- New agent: issue_verification_agent with Workflow-based state machine
- New sub-agent: testing_analyst for AI-driven test result analysis
- New MCP tools: errata (GetErratum, GetErratumBuildNvr),
  testing_farm (GetTestingFarmRequest, ReproduceTestingFarmRequest),
  gitlab (SearchGitlabProjectMrs), jira (UpdateJiraComment,
  AddJiraAttachments, GetJiraAttachment)
- New unprivileged tools: read_logfile, read_readme, search_resultsdb,
  analyze_ewa_testrun, read_attachment
- Supporting modules: baseline_tests, compare_xunit, qe_data
- Shared types added to ymir/common/models.py
- New skill definition for issue_verification

The supervisor's ErratumHandler remains unchanged for a follow-up.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: martinky82

README-agents.md

@@ -11,6 +11,7 @@ Three agents process tasks through Redis queues:
 - **Triage Agent**: Analyzes JIRA issues and determines resolution path. It uses title, description, fields, and comments to find out root cause of the issue. It can ask for clarification, create tasks for other agents or may take no action if not needed.
 - **Rebase Agent**: Updates packages to newer upstream versions. A Rebase is only to be chosen when the issue explicitly instructs you to "rebase" or "update". It looks for upstream references that are linked, attached and present in the description or comments in the issue.
 - **Backport Agent**: Applies specific fixes/patches to packages. It looks for patches that are linked, attached and present in the description or comments in the issue. It tries to apply the patch and resolve any conflicts that may arise during the backport process.
+- **Issue Verification Agent**: Manages the post-fix lifecycle of a JIRA issue — from merged MR through errata creation, testing analysis, and status transitions to RELEASE_PENDING. Migrated from the supervisor's `IssueHandler`.
 
 
 ## Dry run mode

README-supervisor.md

@@ -38,6 +38,10 @@ This will have to be repeated when the ticket expires. (Exporting the kcm-cache
 socket to the container seems better, but I wasn't able to get it to work due
 to various difficult to work around permission issues.)
 
+## Note on IssueHandler migration
+
+The `IssueHandler` has been migrated to `ymir/agents/issue_verification_agent.py` as part of the BeeAI Framework refactoring. The supervisor still handles errata processing via `ErratumHandler`. For processing individual issues, prefer using the new agent directly (see README-agents.md).
+
 ## Processing a single issue or erratum
 
 To process a single issue or erratum, you can run:

agents_as_skills/README.md

@@ -1,5 +1,14 @@
 # These files are skills built from agents
 
+## Available skills
+
+- `backport/` — Backport fix agent
+- `preliminary_testing/` — Preliminary testing agent
+- `rebase/` — Rebase fix agent
+- `rebuild/` — Rebuild agent
+- `triage/` — Triage agent
+- `issue_verification/` — Issue verification agent (post-fix lifecycle management)
+
 ## How to build
 
 ```bash

agents_as_skills/issue_verification/SKILL.md

@@ -0,0 +1,74 @@
+---
+description: >
+  Runs the Issue Verification workflow for a JIRA issue, managing the lifecycle
+  from merged MR through errata creation, testing analysis, and status transitions.
+arguments:
+  jira_issue:
+    description: "JIRA issue key (e.g. RHEL-12345)"
+    required: true
+  dry_run:
+    description: "If true, don't make any changes to JIRA or other systems"
+    required: false
+    default: "false"
+  ignore_needs_attention:
+    description: "If true, process the issue even if it has ymir_needs_attention label"
+    required: false
+    default: "false"
+---
+
+# Issue Verification Agent
+
+Manages the lifecycle of a JIRA issue after a fix has been backported or rebased.
+
+## MCP Tools Used
+
+From the privileged gateway (`MCP_GATEWAY_URL`):
+
+- `get_jira_details` — Fetch full JIRA issue details
+- `edit_jira_labels` — Add/remove labels
+- `add_jira_comment` — Add comments
+- `change_jira_status` — Transition issue status
+- `update_jira_comment` — Update existing comments
+- `add_jira_attachments` — Add attachments to issues
+- `search_gitlab_project_mrs` — Search for merge requests
+- `get_erratum` — Get erratum details
+- `get_erratum_build_nvr` — Get build NVR from erratum
+- `get_testing_farm_request` — Get Testing Farm request status
+- `reproduce_testing_farm_request` — Reproduce a test run with different build
+
+## Workflow Steps
+
+1. **fetch_and_validate_issue** — Fetch JIRA issue, check labels/components
+2. **check_errata_status** — Branch based on whether errata link exists
+3. **run_before_errata** — Check for merged MRs, wait for errata creation
+4. **run_after_errata** — Validate preliminary testing, test coverage, move status
+5. **analyze_testing** — Call testing_analyst sub-agent for test result analysis (INTEGRATION status)
+6. **check_reproduction** — Check status of baseline test reproduction
+
+## Output Schema
+
+```json
+{
+  "status": "string — description of what happened",
+  "reschedule_in": "float — seconds until next check (-1 means don't reschedule)"
+}
+```
+
+## How to Run
+
+```python
+from ymir.agents.issue_verification_agent import run_issue_verification
+
+result = await run_issue_verification(
+    jira_issue="RHEL-12345",
+    dry_run=True,
+    ignore_needs_attention=False,
+)
+print(result.status, result.reschedule_in)
+```
+
+Or standalone:
+
+```bash
+JIRA_ISSUE=RHEL-12345 DRY_RUN=true python -m ymir.agents.issue_verification_agent
+```